/irc-logs / w3c / #css / 2008-08-18 / end

Options:

# Session Start: Mon Aug 18 00:00:00 2008
# Session Ident: #css
# [01:14] * Quits: dbaron (dbaron@82.71.8.145) (Ping timeout)
# [09:06] <Bert> Krijnh, didn't you stop logging in order to patch the software? In particular to allow people to make unlogged remarks, via /me or [off]?
# [09:58] * Joins: dbaron (dbaron@82.71.8.145)
# [10:16] * Joins: Bert_ (bbos@128.30.52.28)
# [10:19] * Quits: Bert_ (bbos@128.30.52.28) (Quit: leaving)
# [11:07] * dbaron gets ready to head out to see some sights
# [11:13] * Quits: dbaron (dbaron@82.71.8.145) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [11:50] * Joins: anne (annevk@77.163.243.203)
# [13:12] * Joins: myakura (myakura@118.8.102.216)
# [13:29] <anne> just to be clear, I'll be in Cambridge too even though http://www.w3.org/Style/Group/2008/Cambridge.html suggests otherwise
# [13:29] <anne> my e-mails don't seem to get through
# [13:44] <Bert> Your e-mail doesn't get through? To the mailing list, you mean?
# [13:51] <anne> to Alex
# [14:37] <fantasai> Bert: that was a suggestion, but not a requirement
# [14:37] <fantasai> Bert, krijnh: See minutes at http://lists.w3.org/Archives/Public/www-style/2008Apr/0456.html
# [14:38] <fantasai> Bert: we resolved that ideally W3C would host the logs and ideally there would be a way to not log a statement, but that for now it's ok for krijnh to log the channel
# [14:38] <fantasai> As for not logging /me
# [14:38] <fantasai> that wouldn't work for me
# [14:38] <fantasai> very well
# [14:38] <fantasai> because I use it all the time by habit
# [14:47] <anne> yeah, not logging /me is misfeature of the W3C bot
# [15:15] * Quits: myakura (myakura@118.8.102.216) (Ping timeout)
# [15:16] * Joins: myakura (myakura@118.8.102.216)
# [15:28] <fantasai> Bert, for colors
# [15:28] <fantasai> Bert: should the Computed value be "as specified" or "the computed color" ?
# [15:33] <Bert> What is the difference?
# [15:33] <fantasai> CSS2.1 says 'as specified'
# [15:33] <fantasai> CSS3 says something long and complicated
# [15:34] <fantasai> http://www.w3.org/TR/css3-color/#foreground
# [15:36] <fantasai> Also, I'm thinking we should change all instances of <uri> to <image> and define <image> as <uri>. That way CSS3 Values and Units can hook into the definition of <image> to define e.g. image slices or gradients or whatever else we decide to add
# [15:36] <Bert> I don't think it matters for CSS itself. We can leave it to Anne to define some canocalization for the DOM, but I'd do the easy thing and say "as specified" in CSS 2.1 and Color
# [15:36] <fantasai> ok
# [15:37] <fantasai> you mean CSS 2.1 and Backgrounds and Borders
# [15:37] <fantasai> I think saying "computed color" rather than as-specified would cover both
# [15:37] <fantasai> actually
# [15:37] <fantasai> think that would work?
# [15:38] <Bert> But how do you compute a color? There are no relative colors, are there?
# [15:38] <fantasai> currentColor
# [15:38] <Bert> Apart from currencolor.
# [15:39] <fantasai> well, that's still one :)
# [15:39] <fantasai> can't just write 'as specified' there, right
# [15:39] <anne> for the DOM it's either the #rrggbb format or rgba(...) in case a != 1
# [15:39] <fantasai> that's what css3-color says
# [15:39] <fantasai> iiric
# [15:39] <anne> (or the canonical name of the system color, I suppose)
# [15:42] <Bert> I don't really care whether you call it "as specified, except for currentcolor" or "see text" and then explain essentially the same thing in the text. Whatever makes the spec shorter is better, I think.
# [15:43] <fantasai> then I'll use "the computed color"
# [15:43] <Bert> And about <image>: where do you want to change that?
# [15:43] <fantasai> everywhere we use <uri> in this spec
# [15:43] <fantasai> because they all represent images
# [15:45] <Bert> Yes, that sounds reasonable.
# [15:46] <Bert> But hwo about the timing/dependency? We're not publishing Values at the same time, are we?
# [15:46] <fantasai> no
# [15:46] <fantasai> so here we define <image> as <uri>
# [15:47] <fantasai> and when Values gets published it can redefine <image>
# [15:47] <fantasai> or rather, expand its definition
# [15:47] <fantasai> like it does with <length>
# [15:48] <fantasai> another thing..
# [15:48] <fantasai> for border-break
# [15:48] <fantasai> I think we should have the current definition plus two keywords, "open" and "close"
# [15:48] <fantasai> where "close" makes each box as if it were independent
# [15:48] <fantasai> and "open" is what we do currently
# [15:49] <fantasai> and then the current definition lets you specify a different border if you really want a different one
# [15:49] * fantasai notes that the initial value isn't a valid value atm
# [15:50] <Bert> Isn't open the same as none?
# [15:51] <fantasai> yeah, I'd replace 'none' with 'open'
# [15:51] <fantasai> 'none' isn't actually defined atm
# [15:51] <fantasai> or rather, it's defined
# [15:51] <fantasai> but not in the list of valid values :)
# [15:51] <Bert> 'None' is a <border-style>
# [15:51] <fantasai> ah
# [15:51] <fantasai> true
# [15:51] * fantasai forgot about that
# [15:51] <fantasai> in that case we'd just need a keyword for close
# [15:52] <Bert> 'Close' might be convenient, indeed.
# [15:53] <fantasai> and 'close' would cause a image border to wrap around each piece of the box
# [15:53] <fantasai> instead of around the unbroken box
# [15:53] <Bert> Good point. So more than just convenient :-)
# [15:53] <fantasai> :)
# [15:55] * fantasai files an issue
# [15:55] <Bert> Hmm, the text currently says that image gets used always. That needs to change then.
# [15:55] <fantasai> yeah
# [15:55] <fantasai> I don't think it makes much sense even
# [15:56] <Bert> It's certainly not very intuitive.
# [15:57] <anne> is any of you going through Heathrow tomorrow?
# [15:58] <fantasai> I'm in London already
# [15:58] <Bert> No, Stansted.
# [15:58] <anne> I see
# [15:58] <anne> oh well, I'll figure out how this bus things works tomorrow as the online thing doesn't work
# [15:58] <fantasai> heh
# [15:59] <fantasai> you can take the train from heathrow to King's Cross
# [15:59] <fantasai> I'd suggest doing that
# [15:59] <fantasai> there's an express from heathrow to downtown
# [15:59] <fantasai> and then you take the Underground from there
# [16:07] <anne> if the national express bus doesn't work out I'll do that
# [16:28] <fantasai> Bert: you might want to tweak dev.w3.org/csswg/default.css to style <code> differently
# [16:29] <fantasai> Bert: add quotes and change the color or whatever you want to do with it
# [16:29] <fantasai> or maybe only add quotes inside a <p> or <dd> or <li>
# [16:30] <Bert> I'll check.
# [16:31] <Bert> "no bib entry found for CSS3Color"
# [16:31] <fantasai> hm
# [16:32] * fantasai fixes errors
# [16:41] <Bert> Hmm, <code> </code> is much longer to type then '' '' :-(
# [16:41] <fantasai> heh
# [16:41] <fantasai> you can write '' '' if you want, I'll fix it before CR ;)
# [16:42] <Bert> Well, the post-processor does that. I replaces '' '' by <code> </code>.
# [16:42] <fantasai> I was using '' for awhile, but it hasn't been post-processing into anything lately
# [16:42] <Bert> Or actually by <code class=css>
# [16:42] <Bert> Then there is a bug...
# [16:42] <fantasai> ah
# [16:43] <fantasai> I've noticed that <code> also does some spiffy auto-linking
# [16:45] <Bert> '' '' seems to work...
# [16:45] <fantasai> not for me
# [16:46] <fantasai> http://dev.w3.org/csswg/css3-background/#the-background-size
# [16:48] <Bert> That's two double quotes, not four apostophes :-)
# [16:49] <fantasai> what is it supposed to be?
# [16:49] <fantasai> ""
# [16:49] <fantasai> or ''''
# [16:49] <fantasai> ?
# [16:49] <fantasai> because '''' is not woking
# [16:49] <fantasai> working
# [16:49] <fantasai> and it wouldn't make sense for "" to convert to <code>
# [16:49] <Bert> The latter.
# [16:49] <fantasai> right, so those are '''' pairs
# [16:50] * Bert doing cvs checkout to see your code.
# [16:50] <fantasai> and they are not getting converted to anything
# [16:50] <fantasai> they look like " in the finished version because of the font
# [16:51] <Bert> Hmm, works for me...
# [16:52] <fantasai> no fair!
# [16:53] * fantasai checks her script and tries setting the group to CSS
# [16:53] <Bert> Ah yes, that does make a difference.
# [16:54] <Bert> curl -F group=CSS ...
# [16:54] * Quits: myakura (myakura@118.8.102.216) (Quit: Leaving...)
# [16:54] <fantasai> ahhh
# [16:54] <fantasai> much better :)
# [16:55] <fantasai> so now the question is
# [16:56] <fantasai> can it do <code>'s magickal auto-linking?
# [16:58] <Bert> Yes, the cross-references are done after the quote expansion.
# [16:59] <fantasai> so should I use '' for both properties and values or just values?
# [17:00] <Bert> single for properties and double for values (or anything that's not a property, such as a declaration or a selector).
# [17:00] <fantasai> will the single quotes <code> and cross-link?
# [17:01] <Bert> They will link to a DFN, if there is any.
# [17:03] <Bert> (The Makefile under css3-src has a "make test-propuse" target that checks if all single-quoted things are indeed properties. The postprocessor doesn't have that check.)
# [17:05] <Bert> The single quotes translate to <span class=property>, but if you think CODE is better than SPAN, that is easy enough to change.
# [17:10] <fantasai> yes, I think we should make that change
# [17:15] <Bert> Done. And I added .property {font-family: inherit}' to default.css
# [17:16] <fantasai> ok
# [17:16] <fantasai> I don't mind it being monospace
# [17:16] <fantasai> but properties and values should be consistent about that imo
# [17:16] <fantasai> BTW, did you notice I changed the markup in the property tables?
# [17:16] <fantasai> I changed them to use <th> instead of <td><em>
# [17:17] <Bert> No, I just saw a blue border...
# [17:17] <Bert> Oh, that table.
# [17:18] <Bert> I don't know what's the best semantic mark-up.
# [17:18] <Bert> But that probably means it's not wrong either.
# [17:18] <fantasai> <th> is definitely better than <td><em>
# [17:18] <fantasai> here
# [17:19] <Bert> Anyway, the style sheet was already changed to handle TH there.
# [17:19] <fantasai> yes, IIRC I did that :)
# [17:19] <fantasai> to cope with other specs I've recoded ;)
# [17:19] <Bert> Probably :-)
# [17:27] <fantasai> can we use generated content instead of the post-processor for the quotes?
# [17:28] <Bert> Hmm, for me they are semantic. Anything in single quotes is CSS code, whether the document is styled or not,
# [17:29] <fantasai> it's already in <code>
# [17:29] <fantasai> that's the semantics
# [17:29] <fantasai> quotation marks is styling
# [17:29] <fantasai> you could use monospace and a border instead, e.g.
# [17:29] <Bert> The <code> is not visible, unless you do view-source.
# [17:29] <fantasai> neither are the paragraph breaks
# [17:30] <fantasai> or the distinction between <h3> and <h4>
# [17:30] <fantasai> these are all controlled by the style sheet
# [17:30] <fantasai> and it's not even <span> we are using here
# [17:30] <fantasai> it's <code>
# [17:30] <fantasai> which has a well-defined meaning in HTML
# [17:30] <fantasai> as well as a default styling that distinguishes it from surrounding text
# [17:31] <Bert> Yes, but there is a default style sheet for HTML, so H3, H4 and P are visible.
# [17:31] <fantasai> and so that <code> looks different from regular paragraph text
# [17:32] <fantasai> that is also part of the default style sheet for HTML
# [17:32] <Bert> But CODE doesn't mean "property," it means "computer code." The quote marks make it into a property.
# [17:32] <fantasai> um
# [17:33] <fantasai> I don't think the quote marks make it into a property
# [17:33] <Bert> They have for the past 12 years or so...
# [17:34] <fantasai> they say "this is a particular bit of text"
# [17:34] <fantasai> in the CSS specs this is typically indicicating some CSS syntax
# [17:34] <fantasai> the same quote marks in other specs indicate other types of syntax
# [17:34] <fantasai> but really that distinction is already marked up in the <code> tags
# [17:34] <fantasai> the quote marks are extra
# [17:34] <fantasai> and not necessary in the source
# [17:35] <fantasai> they are necessary in your styling because you took out the monospace distinction that is there by default
# [17:35] <Bert> The CSS specs have been pretty consistent in their use of single and double quotes.
# [17:36] <fantasai> CSS2.1 uses ' to mark CSS syntax
# [17:36] <fantasai> in general
# [17:36] <fantasai> it does not use <code>
# [17:36] <fantasai> CSS3 specs should use <code>
# [17:36] <fantasai> that is more correct
# [17:37] <fantasai> they do not to use both <code> and single quotes
# [17:37] <fantasai> that is excessive
# [17:37] <fantasai> imo
# [17:37] <fantasai> s/not to/not need to/
# [17:37] <Bert> Then I prefer the human mark-up (quotes) to the computer mark-up (tags)
# [17:38] <fantasai> you're using computer mark-up for paragraphs and headings
# [17:38] <fantasai> why not for computer code?
# [17:38] <Bert> Yes, having both is convenient for geeks like me, who write perl scripts "-)
# [17:38] * Joins: sylvaing (sylvaing@131.107.0.102)
# [17:39] <fantasai> My opinion is that <code> + generated content would be best.
# [17:39] <fantasai> I think that is most in-line with how HTML is meant to be sued
# [17:39] <fantasai> s/sued/used/
# [17:39] <Bert> But it is very convenient, not just in the spec, but also in e-mail, to consistemtly use single quotes to mean "CSS code".
# [17:39] <fantasai> heh
# [17:39] <fantasai> sure
# [17:39] <fantasai> do that in the pre-processed version
# [17:39] <fantasai> for the spec
# [17:40] <fantasai> and always in email
# [17:40] <fantasai> the post-processed version of the spec will display the quotes using GC
# [17:40] <fantasai> but won't need to have it in the spec itself
# [17:41] * fantasai personally finds the proper single quotes noisy and wants the option to turn them off :P
# [17:41] <fantasai> can't turn them off if they're in the source
# [17:41] <fantasai> ;)
# [17:41] <Bert> That's too much difference between the styled and unstyled documents in my mind. I want the styled version to look nicer, not have different content.
# [17:42] <fantasai> it is the same content
# [17:42] <fantasai> it just looks different
# [17:42] <fantasai> monospace vs. quotes
# [17:42] <fantasai> not indistinguishable vs. quotes
# [17:43] * fantasai really likes the styling in hixie's specs, color + monospace for code snippets
# [17:43] <fantasai> although I probably wouldn't have used bright red
# [17:44] <Bert> But monospace can also mean "HTML code" or "DOM API."
# [17:44] <fantasai> so can quotes
# [17:44] <fantasai> they can also mean many other things
# [17:44] <Bert> (And, as I said, I find monspace difficult to read when mixed with proportional.)
# [17:44] <fantasai> quotes do not mean CSS. They're just a means to distinguish this snippet from the surrounding text.
# [17:44] <fantasai> Like monospace
# [17:44] <fantasai> like color
# [17:44] <fantasai> like underlining
# [17:44] <fantasai> like italics
# [17:44] <fantasai> it does not matter how you style it
# [17:45] <fantasai> as long as you style it differently
# [17:45] <fantasai> and consistetnly
# [17:45] <Bert> Not in CSS specs. When talking about CSS, single quotes mean CSS code.
# [17:46] <fantasai> because that is the styling convention you chose
# [17:46] <fantasai> it is only a styling convention
# [17:46] <Bert> It happened, nobody chose it.
# [17:47] <fantasai> I'm not saying to change it, I'm saying to change how it is coded
# [17:47] <fantasai> so that the styling is kept
# [17:47] <fantasai> but the code is consistent with HTML
# [17:47] <fantasai> and how HTML defines its own semantics
# [17:48] <fantasai> anwyay
# [17:48] <fantasai> *anyway
# [17:48] * fantasai can't type today
# [19:05] * Joins: dbaron (dbaron@82.71.8.145)
# [19:24] * Quits: sylvaing (sylvaing@131.107.0.102) (Quit: sylvaing)
# [19:38] * Quits: arronei (arronei@131.107.0.106) (Ping timeout)
# [19:44] * Joins: arronei (arronei@131.107.0.106)
# [20:21] * Quits: dbaron (dbaron@82.71.8.145) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [20:24] * Joins: sylvaing (sylvaing@131.107.0.77)
# [20:41] * Quits: arronei (arronei@131.107.0.106) (Ping timeout)
# [20:46] * Joins: arronei (arronei@131.107.0.106)
# [21:18] <anne> sylvaing, hey, hope my reply was not too late, have been away for a while
# [21:25] * Joins: glazou (daniel@89.159.122.239)
# [21:26] * Quits: glazou (daniel@89.159.122.239) (Quit: glazou)
# [21:33] * Joins: Cedric (chatzilla@90.36.176.190)
# [21:34] * Parts: Cedric (chatzilla@90.36.176.190)
# [21:38] * Joins: glazou (chatzilla@89.159.122.239)
# [21:38] <glazou> hi
# [21:57] * Quits: glazou (chatzilla@89.159.122.239) (Connection timed out)
# [23:57] * Joins: dbaron (dbaron@82.71.8.145)
# Session Close: Tue Aug 19 00:00:00 2008

The end :)